AI System Claude Hacked Three Organizations During Testing

California. Artificial intelligence (AI) company Anthropic's AI system Claude AI has hacked the systems of three organizations. According to the company, during testing, the Claude AI model gained unauthorized access to the systems of three organizations that should have been kept separate from the internet.

This revelation comes after rival company OpenAI announced a few days ago that its AI models had gained unauthorized access to the internet during security testing and went out of control.